Roubini Says AI Will Force Universal Basic Income or Socialism

Nouriel Roubini, the economist nicknamed “Dr. Doom” for calling the 2008 crash, says AI will force governments to choose between universal basic income and ownership stakes in Big Tech. He made the case Friday on Bloomberg Television, and he insists it is the optimistic version of events.

The prediction did not start as a question about robots. Bloomberg anchors had asked him how to fix Social Security, whose trust fund is projected to run dry in 2032. His answer skipped past retirement-age fixes entirely.

A Retirement Question Becomes a Theory of Everything

Roubini was asked how policymakers might shore up Social Security’s finances. Raising the retirement age, a fix that keeps resurfacing in Washington, was on the table.

He said it would not be enough. A large share of the workforce, he argued, will be replaced by AI and robots within the next 20 to 25 years, making a higher retirement age beside the point for millions of workers.

“Eventually, we need some form of universal basic income for everybody while they work and once they retire,” Roubini said during the Friday interview on Bloomberg Television. “We’re already on the way.”

The Growth Math Behind the Optimism

Roubini’s case rests on a specific growth forecast, not a vague hunch about robots taking over jobs. He calls the AI boom the most consequential wave of tech innovation in human history, one that eventually produces artificial general intelligence, meaning machines that match or beat human cognitive ability.

That shift, in his telling, changes the global growth rate itself.

• Today through the late 2020s: global growth of roughly 2% to 4% a year.
• By 2040: growth accelerating toward 6% a year.
• By 2050: growth reaching 10% a year, with machines doing most of the work.

Once growth hits that level, Roubini argues, government has something new to tax: the profits flowing to whoever owns the machines. That is where the redistribution math begins.

Ex-Post Versus Ex-Ante, Two Roads to the Same Place

Roubini frames the choice ahead in an economist’s shorthand: distribution can happen after profits are earned, or before.

We’ll have either ex-post distribution, or we’ll have it ex-ante. Ex-ante means some form of socialism.

He used ex-post as shorthand for universal basic income, cash paid out after profits are taxed. Ex-ante describes the opposite: government holding a stake in the companies from the start. “Essentially, the government is going to take over some fraction of the big tech firms,” he said.

Roubini is not the only one describing a version of that plan. Anthropic, the AI company behind Claude, published its own policy paper last month calling for what it termed “universal pre-distributive capital accounts,” with priority given to workers displaced by AI, according to a Time report on the AI ownership talks. Roubini pointed to this exact kind of arrangement in his own interview, noting that AI companies are already willing to hand equity to the government as a way to let the public share in AI’s upside.

Washington Already Owns a Piece of the Future

That is not hypothetical. OpenAI chief executive Sam Altman has proposed giving the U.S. government a 5% equity stake in the ChatGPT maker, according to a Financial Times report cited by CNBC. Under the pitch, other major AI developers would hand over similar stakes through a single government-linked fund.

It would not be Washington’s first stake in a private tech company. Here is how the arrangements compare.

The Intel arrangement is the precedent already locked in. President Trump has defended the broader approach, telling reporters he thinks it is “very American” for Washington to hold equity in private firms and that the concept would make ordinary citizens “partners in this revolution,” according to CNN’s report on the talks, which noted the proposed AI vehicle is modeled on Alaska’s oil-wealth dividend fund. Nothing is final. Any formal deal would likely need an act of Congress, and it is unclear whether Anthropic, Google or Meta would sign on to a plan Altman floated largely on OpenAI’s behalf.

The Social Security Clock That Started This Conversation

The question that opened the Bloomberg interview is not abstract. Social Security’s Old-Age and Survivors Insurance trust fund is projected to run out in the fourth quarter of 2032, one quarter sooner than last year’s estimate, according to the Social Security trustees’ 2026 summary report.

Once that reserve is empty, the fund could still pay 78% of scheduled benefits from ongoing payroll tax revenue, not zero, but a real cut. The nonpartisan Committee for a Responsible Federal Budget estimates the typical retiree would lose roughly $500 a month if Congress does not act first.

Roubini’s point was that raising the retirement age assumes people can keep working longer. His forecast says a shrinking share of them will have that option at all, regardless of age, because the jobs will not exist in the same numbers.

Does Giving People Free Cash Change How Much They Work?

Yes, slightly, according to the largest real-world test of the idea in the U.S. so far. OpenResearch, the nonprofit Altman chairs, ran a three-year, $1,000-a-month cash pilot in Illinois and Texas, and the 2024 results are the closest thing to hard data on what Roubini and Altman are both describing.

The payments amounted to a 40% income boost for the 1,000 recipients, compared with a 2,000-person control group that received $50 a month, according to CBS News’s review of the study.

• Recipients worked about 1.3 fewer hours a week than the control group, and were 10% more likely to be actively job hunting.
• Researchers found no improvement in job quality among recipients who kept working. “We don’t find any effects on the quality of employment whatsoever,” one of the study’s researchers, Vivalt, told CBS News.
• Hospitalizations among recipients rose 26% compared with the control group, and emergency department visits rose about 10%.
• Spending on essentials and on supporting family and friends rose, and recipients did not drop out of the workforce, according to Bloomberg’s coverage of the OpenResearch findings.

That data does not settle whether basic income works. It shows the version tested so far looks less like Roubini’s vision of machines doing all the work, and more like a modest cushion rather than a full paycheck replacement.

Altman and Musk Are Betting on the Same Off-Ramp

Altman has pushed a version of this idea for years without fully committing to it. He has donated to former presidential candidate Andrew Yang, whose 2020 campaign centered on a monthly “freedom dividend,” and for years he chaired OpenResearch while giving lead researcher Elizabeth Rhodes full independence over the pilot, according to Newsweek’s reporting. OpenAI’s actual policy pitch to Washington, though, is an equity stake, not a guaranteed check.

Elon Musk has described a friendlier version of the same endpoint, arguing that work becomes optional rather than mandatory once AI and robotics mature.

“You can grow your own vegetables in your garden or you could go to the store and buy vegetables,” Musk said on the December episode of the People by WTF podcast. “It’s much harder to grow your own vegetables. But some people like to grow their vegetables, which is fine. But it’ll be optional, in that way, is my prediction.”

Roubini’s optimism rests on the same assumption Musk is making, that growth arrives before the social contract breaks. Nothing about that timeline is guaranteed. No company has signed on to the ownership structure that would fund it, either.